Not sure how dynamic analysis would be really helpful here, as hardware watchpoints on the relevant peripheral register addresses apparently don't trigger (I assume they only work for ram) and the debugger doesn't survive the watchdog reset:
Looking at the source, I see that Arduino-pico is running its OTA handler before going in the user application, so it must be something in there. I only see it messing with the clocks peripheral though, so I guess I'll have to dig a little deeper there.
Code:
(gdb) cContinuing.[...watchdog reset happens]Failed to select multidrop rp2040.dap0Could not read registers; remote failure reply 'E0E'(gdb)Statistics: Posted by kbraden — Sun Dec 15, 2024 1:54 pm